Description

UNM is a high quality cigar butt.  It has a 12% FCF yield (7x 2010 EPS) and it is doing large buybacks at 0.8x tangible book value.  Book value is solid and growing.  There is 50% upside potential over the next year from growth in book value per share and from multiple expansion off of the current trough level.  UNM is the dominant company in its industry.  Management is excellent.  They are disciplined on pricing, choosing profits over growth, they are conservative with the investment portfolio, and they are returning capital to shareholders.  The downside is limited due to stable book value growth, highly accretive buybacks, and the current trough valuation.  Competitors have recently stumbled, and the pricing environment is set to improve, aiding growth and profitability.

Description

UNM is the largest disability insurer in the US with approximately 20% market share and it has been the dominant provider for over 10 years.  It is also a leading provider of group life insurance and supplemental insurance such as accident and critical illness insurance.  Its products are sold primarily to companies as employee benefits.  Supplemental products are offered on a voluntary basis to employees through payroll deductions.  UNM reports in four main segments:

1.      Unum US, 57% of pre-tax, which includes group disability (#1 market share), group life insurance (#7 market share), and supplemental products (#3 market share).

2.      Colonial, 21% of pre-tax, is a separate company within UNM which is like a mini-AFLAC.  It is #2 behind AFLAC in selling voluntary benefits such as life, accident, sickness, cancer and critical illness products.  It is considered a crown jewel with ~20% ROEs and it is the fastest growth segment of the company.  It is distinct from Unum US because it has its own sales force which targets small blue-collar companies that do not offer many benefits.

3.      UNM UK, 19% of pre-tax, where it is the largest disability insurer in the UK with 50% market share. 

4.      Closed Block, 3% of pre-tax. This is a discrete block of policies written in the early 1990's primarily to doctors.  The product was poorly underwritten, allowing customers to easily claim a disability.  This block is segregated and partially reinsured, and it will run-off over 20 years.  It is well-seasoned and earnings are stable, but it ties up a lot of capital and earns a 1% ROE. 

History

UNM's history is important because it helps to explain and to highlight the quality and conservatism of the current management team.  UNM was formed by the merger of two companies in 1999.  The period from 1999-2003 is marked by several major problems.  The old CEO stretched for growth by expanding in the "large case market" (i.e. F500 companies) because it is easier to show top line growth by landing big premium accounts.  Large case business is less profitable, in general, and they mis-priced it to win accounts. 

Compounding the problem, UNM fraudulently denied disability claims in order to enhance profits on the mis-priced business.  This lea to lawsuits and eventual regulatory settlements, and UNM completely revamped its process for handling claims.  On top of all this, UNM had a large exposure to high yield debt going into the 2002 default cycle which caused large losses and forced it to raise capital. 

Since 2003, when the current CEO took over, UNM has been making broad operational improvements including: (i) raising prices and turning away less-profitable business, (ii) shifting from large accounts to more profitable small and mid-size accounts, (iii) cleaning up the investment portfolio to the point where it is the safest in the life insurance industry, and (iv) building excess capital.  (See below for more information on management and operational improvements.) 

Financial Characteristics

UNM's business is fairly stable and predictable.  Revenue is largely recurring with 85-90% renewal rates in most product lines.  Incidence of new insurance claims and claim recoveries are predictable.  The "operating" ROE (i.e. ex-investment portfolio gains/losses) has been 11% for the last three years, up from 9% in 2005/06.  Tangible book value per share has grown consistently every quarter for the last three years, including through the financial crisis.  Tangible book value per share is currently growing at around 3.0% each quarter.

Premium revenue is tied to corporate payrolls and salary increases, and premiums are down 1% YTD.  Longer-term, UNM should grow the top-line at a GDP-like rate.  In 2011, I expect it will grow revenue and net income in a range of 1-3%, and improve from there, assuming a combination of very weak payroll growth and salary increases, and some price increases.

UNM is levered 6.5x-to-1 with a $45bn investment portfolio.  The portfolio is perhaps the most conservative in the insurance industry.   It is mostly A-rated corporate bonds, but what really stands out is that there is essentially no exposure to private mortgage bonds (sub-prime, Alt-A, etc.), commercial real estate, or asset backed securities.  These categories account for 30% of the average life company's investment portfolio.  UNM realized only 80bps of losses cumulatively on its portfolio in 2008 and 2009.  Contrast this to Prudential and Hartford, two of the largest life insurers, which lost 230bps and 430bps, respectively.

Management

The management team is excellent:

1.      They have focused on pricing and profitability instead of growth.  In the core group disability business, they have shrunk premium by 2-5% in each of the last 5 years, and at the same time the benefit ratio (i.e. claims expense/premium revenue, also called the loss ratio) has dropped from 95% to 85%.  This is in stark contrast to competitors as discussed below.  Management has expressed continued commitment to hold the line on prices. 

2.      They have been extremely conservative with the investment portfolio.

3.      They have built significant excess capital which they are committed to returning via buybacks.  They were the first life insurance company to announce a buyback since the financial crisis.  (So far one other, AMP, has followed them.)  They announced a $500mm buyback in May.  They bought back $130 mm in Q2, and they indicated they will do the rest in the second half of 2010, which represents 5% of their market cap.  Management has indicated that buybacks are a key part of their long-term strategic plan, and they believe they can take the ROE from 11-12% to 12-14% via buybacks.  I expect they will execute aggressively on a new authorization once the current authorization is utilized.

Valuation

(Reference the tables below)

1.      UNM is valued at 0.83x current tangible BV (all book value figures exclude AOCI which adds another $1.30 to book value per share), 0.79x 2010E BV and 0.70x 2011E BV.  Tangible BV per share is growing about 13% per year which reflects an 11% ROE plus the accretion from buying back stock below BV. 

2.      The P/E is 7.7x/6.9x on 2010/2011 EPS.  The distributable FCF yield is 11.0%/12.4% in 2010/2011.  If you adjust for $1.50 per share of excess capital at the holding company, the P/E is 7.1x/6.4x and the FCF yield is 11.9%/13.4%.  There is an additional $1.50 per share of excess capital within the insurance subsidiaries, but I don't adjust for it because management has indicated they want to run with higher capital levels for now.

3.      I think there is 50% upside over one year.  UNM has traded on average at 1.0-1.1x book value and 10x forward EPS over the last five years, including the depressed valuations during the financial crisis.  I use 1.1x book value in my upside case, which is consistent with an insurer earning an 11% ROE.  This is also consistent with a P/E of 10x.  The 50% upside over one year reflects 13% from BV growth, 32% from multiple expansion, and 2% from the dividend.

4.      I think the return profile is asymmetric.  A) Book value is solid because the investment portfolio and reserves are conservative.  B) Book value per share is growing because of stable earnings, and because most of the FCF is being used right now to buyback stock below BV.  C) The P/B and P/E ratios are at the lowest levels in 5 years except for a 6-month period during the financial crisis.  At a minimum, the returns should be decent just from BV growth and modest multiple expansion from trough levels.

Free Cash Flow

In terms of calculating FCF, I look at it a two ways shown below.  UNM reports statutory earnings as required by insurance regulators, which is more conservative than GAAP earnings.  Stat earnings are commonly thought of as representative of cash flow because customer acquisition costs are expensed as incurred instead of being capitalized and amortized.  I calculate cash flow from the GAAP financials by starting with net income and subtracting capex, net of depreciation, and subtracting capitalized customer acquisition costs (aka deferred acquisition costs, or "DAC").  This ties pretty closely with reported stat earnings.  I also make an adjustment to retain equity in order to maintain the risk-based capital ratio as UNM grows.  I assume statutory equity needs to grow 4%, in-line with what I consider to be the longer-term normalized growth for the business.  However, this is more of a theoretical adjustment since UNM is not currently growing, so I am understating the actual FCF generation.   

 Why is it Mis-Priced?

I think UNM is overlooked and under-appreciated within the life insurance sector, which itself is depressed.  It is easy to overlook UNM because it has a niche specialty with its focus on disability, and because it is small with a market cap of $7bn vs. heavyweights like Metlife, Prudential and AFLAC, which are all over $20bn.  The entire life insurance industry is under a cloud after a few major companies almost went bankrupt during the financial crisis.  The sector trades at 1.0x book value and 8x earnings, which is well below long-term historical valuations of 1.4x book value and 12x earnings. 

The major industry issues include investment portfolio risk and equity market risk, neither of which UNM has, but UNM still trades at a discount to the group.  As noted above, the industry still has substantial exposure to commercial real estate which is not "out-of-the-woods" yet.  In terms of equity market risk, the financial crisis revealed that most of the industry mis-priced equity-linked variable annuity products and provided guarantees on account balances without proper hedges.  Many guarantees are underwater and remain a major risk if equity markets decline.

Finally, I think the sell-side is under-estimating the share-buyback potential.  UNM was the first insurance company to announce a buyback, which it did at the end of May.  Almost every sell-side analyst is assuming fewer share repurchases in 2011 than in 2010 despite the fact that the 2010 buybacks reflect only a partial year, and the fact that UNM will end 2010 with more excess capital than it started.  Looking at the estimates from seven analysts, the average share buyback for 2011 is 14 mm shares while I estimate they could buyback 25 mm shares even assuming a $30 share price.

Current Issues

Pricing Environment

I believe the pricing environment for disability will start to improve because all of UNM's competitors have experienced deteriorating loss ratios in the last 18 months and because UNM's largest competitor, Hartford, recently said it has started to raise prices.  The following exhibit is tells an interesting story.  The take-away is that UNM's competitors have been growing disability premium in the mid-to-high single digits in recent years while UNM has been shrinking mid-single digits.  UNM's benefit ratio has consistently been going down while the benefit ratio at competitors has been on an increasing trend for 18 months.  This indicates that UNM has been more disciplined with pricing than competitors.  I think the natural reaction at competitors will be to raise prices, which at least Hartford has publicly confirmed.  UNM has also said they are cautiously optimistic that pricing will improve.

Low Interest Rates

The prospect of perpetually low interest rates is a topical concern for life insurers.  It is a low risk for UNM especially in the short to medium term for several reasons.

1.      UNM is largely duration matched.  The investment portfolio has a 7.7 year duration and the liability duration is less than one year longer.  A significant duration mis-match is what killed the Japanese insurers 10 years ago.

2.      There is minimal re-investment risk (i.e. rolling over maturing bonds into much lower yielding bonds) because UNM only has to reinvest 6% of its portfolio this year.  This is much lower than the duration of the portfolio would imply because about one-third of the portfolio is associated with the Closed Block which is in run-off and which has minimal reinvestment needs. 

3.      UNM has a significant "reserve margin" of 100 bps which represents the spread between the investment portfolio yield and the rate used to discount reserves.  This is the highest that the margin has ever been, and it is well above UNM's target of 50-60bps.  This shows that management has conservatively set reserves.  (By contrast, UNM's pure-play competitor Stancorp has a 38 bps margin.)  This is significant because UNM has the flexibility to allow the reserve margin to drift down if the investment yield drops.  Without this flexibility, UNM would have to reduce the discount rate for its reserves which would result in a higher current period benefit ratio since the present value of future losses is higher.

4.      In the long term the entire industry would need to raise prices to adjust for the fact that less revenue would come from interest income vs. premium.  It is essentially a commodity industry with low-ROE's, so all the players would need to raise prices in order to continue to earn their cost of capital. 

Catalyst